CA1175571A - Method and apparatus for identifying objects - Google Patents

Abstract

METHOD AND APPARATUS FOR IDENTIFYING OBJECTS

ABSTRACT

Disclosed is a method and an apparatus for identi-fying objects, such as articles sold in a store, appearing in a random position and orientation and at random times on an image window. On a surface facing the image window, each object has an identification in the form of a field which comprises on at least one data track contrasting indicia with at least one contrasting line pattern identifying the posi-tion and the orientation of the data track. The track also includes a pluraligh of parallel lines with variable spacing and/or line widths. The image window is scanned line-by-line to generate a vinary video signal which corresponds to the scanned contrast sequence. The length of the overlapping light and dark intervals of the video signal resulting from the scanning of the line pattern is measured and successively measured interval lengths are compared with each other and a comparison signal having a first amplitude is generated when the two interval lengths which are being compared have a predetermined ratio to each other which conforms to the spacing of corresponding pattern lines. An identification signal is emitted when during each of a number of successive comparison steps, the number of which is determined by the contrasting line pattern, a reference signal having the first amplitude is generated.

Description

1~755`'7'i Method and apparatus for identifYinq objects The invention relates to a method and an apparatus for identifying objects appearing in a random position and orientation and at random times on an image window, and represents an improvement to the sole invention of Klaus Wevelsiep disclosed and claimed in Canadian Patent No.
1,151,299 issued on August 2, 1983. On a surface facing the image window, each object has an identification in the form of a field which comprises on at least one data track contrasting indicia with at least one contrasting line pattern identifying the position and the orientation of the data track. The track also includes a plurality of parallel lines with variable spacing and/or line widths.
The image window is scanned line-by-line and a binary video signal is generated which corresponds to the scanned contrast sequence. In a first step, the image window is scanned from varying angles or directions until a con-trasting line pattern is detected. In a second step, the position and alignment of the data field relative to the image window is determined and in a third step, raster scan is preformed in the direction of the data track to read and decode the indicia present on the data track.
Such method and apparatus are already known. The objects to be identified are, for example, commercial goods, department store articles, or the like which are identified in machine readable form. For this purpose, appropriate identifications are applied to the objects by imprinting thereon with the desired code, for example the OCR code. The uncoded information may comprise indications of quality, size, price, the number of the articles, and so forth.

1175~

It is difficult to machine read such codes since the objects vary in size and since the code is frequently printed on adhesive labels which are applied to varying points on the article. Therefore, it cannot be assumed that the information can be found in a 6pecific location with a fixed orientation and at predetermined time intervals. Thus, the reading of such codes cannot be compared with the reading of punched cards or the like, where a card is available in a precisely defined reading position at precisely fixed times.
In the present case, the exact opposite applies. The data field on the object appears only more or less approximately at a specific place, and the alignment of the data field is to some extent arbitrary.
This type of method and apparatus for the identifi-cation of an object is used, for example, at the cash counters of supermarkets and the like in order to identify the price and/or the number of the articles which a customer wishes to buy and which he has brought to the cash counter for this purpose. The articles, such as boxes of varying shape and size, bottles, cartons, cans, and the like, are then placed individually into the image window with the surface bearing the data field directed toward the image window. The data fields on the various objects thus appear in variable alignment at differing locations within the image window. The data fields also do not appear at the scanning station at fixed time intervals. Thus, the scanning station must be able t~ search for the data field and, once found, must readout the data track signals in the direction of the data tracks of the field. The readout signals can then be 3~ fed to the cash register in the form of electric impulses so that the cash register can print out on the cash receipt the price, the number of the article, its classification, etc.
The data field applied onto the object is provided with a contrasting line pattern or product identification code ("PIC") which is defined ~y a plurality of parallel lines of varying spacing and/or line width. The purpose of the contrasting line pattern is to reliably and clearly distinguish the data field, for example the printed label, ~`

117~571 from other indicia or line patterns which may be present on the object in the vicinity of the data field. Further, within the data field the contrasting line pattern has a given position and orientation which can be used to ascertain the position and orientation of the contrasting line pat-tern -- and thereby of the data tracks -- relative to the scanning angle of the scanning line. This can then be used to subsequently raster scan the field in the direction of the data tracks perpendicularly over the code lines. Thus, the reliable ~dentification of the contrasting line pattern represents an important step in the omnidirectional reading of visibly imprinted data fields.
The German Offenlegungsschrift 2,338,561 discloses a method and an apparatus of the above described type wherein the identification of the contrasting line pattern occurs only when the lines of the pattern are oriented substantially perpendicularly to the scanning direction and the resulting pulse seguence of a video signal generated thereby equals a predetermined pulse sequence which corresponds to the con-trasting line pattern used. In other words, theOffenlegungsschrift discloses a correlation method. It is particularly disadvantageous that the scanning of the image field for locating the contrasting line pattern must proceed in very small angular increments until a scanning line trans-sects the pattern vertically. This results in an undesire-ably long search time. Further, blurred printed edges of the contrasting line pattern may prevent a recognition of the pattern since in such an event the scanned pulse sequence and the resulting video signal may deviate from the stored refer-ence pattern. As a result, either the search process must berepeated or the contrasting line pattern is not recognized at all, leaving the associated data field unread.
In contrast, it is an object of the invention to provide a method and an apparatus of the above described type which make possible a rapid and reliable identification of contrasting line patterns while generally preventing blurred edges of the lines from adversely affecting the readout.

~1755'71 In accordance with the previous invention of Wevelseip mentioned above this is achieved by me~suring the length of the overlapping light and dark intervals of the video signal resulting from a scanning of the line pattern. Successively measured interval lengths are compared with each other and a reference or comparison signal having a first amplitude is generated when the two interval lengths which are being com-pared have a predetermined ratio to each other which conforms to the spacing of corresponding pattern lines. An identifi-cation signal (PIC OUT) is emitted when during each of a numberof successive comparison steps, the number of which is deter-mined by the contrasting line pattern, a reference signal having the first amplitude is generated.
The apparatus has an optoelectronic scanner which outputs binary video signals that correspond to the line-by-line scanned image field and comprises a series of light and dark intervals. The apparatus has a decoder for identi~ying a seanned eontrasting line pattern of a plurality of parallel lines witn varying spaeing and/or line widths whieh eharaeterizes the position and orientation of at least one data track of the data field. The apparatus further ineludes means for aligning the scanner parallel to the data track and for reading the indieia of the seanned data traek.
In aceordanee with said previous invention the apparatus is eharaeterized by a eounting cireuit whieh reeeives the video signals and determines the length of sueeessive, overlapping video signal intervals. Further, the apparatus has at least one referenee table whieh reeeives sueeessively eounted interval lengths in pairs via a gate eireuit and emits a eomparison signal having a first amplitude when two eompared intervals lengths have a given ratio whieh eorresponds to the ratio of the eorres-ponding interval of the eontrasting line pattern. An evaluation eireuit is provided which generates an identi-fieation signal (PIC OUT) when a predetermined number ofreferenee signals with a first amplitude have been generated.

1~7~57~

The advantages of the invention reside particularly in that t~e identification of the contrasting line pattern is dependent only on whether the ratio of successive overlapping interval lengths of the video signal lies within narrow ranges. As a result, the contrasting line pattern is reli-abily identified even when scanning at an oblique angle to lines of the pattern since the ratios of successively mea-sured interval lengths of th video signal are constant and do not vary with the scanning angle. Thus, a single scanning of the contrasting line pattern -- at any desired angle -- so long as all lines of the pattern are crossed, is sufficient ; for a reliable identification of the line pattern. The line pattern can, therefore, be rapidly identified with relatively few scans in which the angular inclination is varied in large increments.
The decoding of the video signal for detecting the contrasting line pattern in relation to the video signal must take place in real-time, that is substantially simultaneously with the generation of th video signal to ensure that the line pattern is identified anywhere within the video signal.
Therefore, the determination of whether the ratio of succes-sively measured interval lengths falls within a given value range, is carried out not by a division but by entering the value of the ratio into a two-dimensional comparison table, sometimes also called a division table. The table emits a reference signal of a given, first amplitude only when two interval lengths which are being compared lie in a predeter-mined field of th~ table where the guotient of the compared interval lengths has a value range which corresponds to that of the corresponding interval length of the searched for line pattern. The time and hardware consuming division process is thereby eliminated and real-time operation with regard to the video signal is maintained.
The comparison table is preferably in the form of a two-dimensional programmed read-only memory (PROM). The various possible discrete values of the first measured inter-val length address the lines of the memory. The various discrete values of the next following measured interval ~175S~l length address the columns of the memory. The expected field is defined so that it encompasses all storage points where the quotient of the interval length associated with the line address fall within a given value range. If a storage point within the expected field is addressed, then the PROM emits a comparison signal of a first amplitude which signals a partial identification of the line pattern.
If, however, the storage point determined by the line column addresses lies outside the expected field then a comparison signal with a second amplitude is emitted to indicate that the quotient of the compared successive interval lengths lies outside the predetermined value range. When a refer-ence signal with a second amplitude appears, the decoder is reset and ready for a new identification and decoding step.
Preferably, two interval lengths which are to be compared are fed into a separate comparison table with its own expected field and examined there with regard to their quotient value. Each individual comparison table preferably comprises as a separate PROM. This simplifies the control logic of the decoder.
In accordance with a preferred embodiment of the invention, eight bit memory positions are provided for the first PROM to compare the first and the second interval lengths as well as for the second PROM to compare the second and third interval lengths, and so forthO Thus, eight com-parison tables, each with its own expected field can be accommodated in the PROM's, the first table being formed for example from the first bit of the memory positions, the second table from the second bit of the memory positions, and so forth. By selectively addressing and selectively reading out the table which is being used, it is possible to identify up to eight different contrasting line patterns with one decoding circuit, thereby enhancing the utilization of the apparatus in accordance with one aspect of the invention.
In accordance with the present joint invention of Wevelsiep and Kastner, a first expected field is provided for identifying the interval length of a certain contrast line pattern measured successively in forward direction. A second expected field is provided for identifying these interval length successively measured in rearward direction.The contrast ~ ~75~

line pattern is scanned in ~forward direction" when the contrast lines are read for example from left to right.
The opposite direction is called the "rearward direction", i.e. in the rearward direciton the contrast lines are scanned opposite to the usual reading direction, i.e. from right to left. If scanning is carried out in forward - direction, and if the addressed lines and columns of the successively measured interval length lie in the first expected field, an identification signal PIC OUT V is emitted which includes the information that the contrast line pattern is identified in forward direction. If how-ever, the contrast line pattern is identified by scanning in rearward direction, a second identification signal PIC OUT R is emitted which is different from first iden-tification signal PIC OUT V. Thus, the first method step to find the contrast line pattern is substantially acceler-ated, i.e. the number of different angle scanning steps is reduced when the contrast line pattern is identified in forward and rearward direction. Time for carrying out one reading cycle is considerably reduced.
Thus, one aspect of the present invention can be expressed as a method for identifying objects appearing at random positions in random orientation, and at random times on an image window and having, on a surface facing the image window, an identification in the form of a field which includes on at least one data track contrasting indicia with at least one contrasting line pattern iden-tifying the position and the orientation of the data track and having a plurality of parallel lines with variable spacing and/or line widths, the image window being scanned line-by-line and a binary video signal being generated which corresponds to the scanned contrast sequence, the image window being, in a first method step, scanned under varying angles until a contrasting line pattern is detected, the position and alignment of the data field relative to the image window being determined in a second ~755'~
- 7a -method step and, in a third method step, a raster scan in the direction of the data track being performed and the indicia present on the data track being read and decoded, the contrasting line pattern being identified by measuring S the length of the overlapping light-dark intervals of the video signal, comparing the successively measured interval lengths, generating a comparison signal having a first amplitude when two interval lengths which are being compared have a predetermined ratio to each other which corresponds to a corresponding spacing in the contrast-ing line pattern, and emitting an identification signal when, during each of a number of successive comparing steps as determined by the contrasting line pattern, the comparison signal is received, the comparing of lS successively measured interval lengths is carried out in pairs in a two-dimensional comparison table within which the possible discrete values of an interval length are assigned to table lines, and the possible discrete values of the next following measured interval length are assigned to table columns, and wherein an expected field is specified encompassing the positions on the table in which the quotient of the two successively measured compared interval lengths falls within a given value range, and including the step of generating the refer-ence having a first amplitude when the compared interval lengths correspond to a table position within the expected field, differing expected fields for corresponding success-ively measured interval lengths of various contrasting line patterns are included in the individual comparison table, characterized in that an individual identification signal for each contrasting line pattern is emitted when identifying the corresponding contrasting line pattern.
Another aspect of the present invention can be ex-pressed as an apparatus for identifying objects appearing at random positions in random orientation, and at random times on an image window and having, on a surface facing 117~57~
- 7b -the image window, an identification in the form of an image field which includes on at least one data track contrasting indicia with at least one contrasting line pattern identifying the position and the orientation of the data track and having a plurality of parallel lines with variable spacing and/or line widths, the image window being scanned line-by-line and a binary video signal being generated which corresponds to the scanned contrast sequence, the image window being, in a first step, scanned under varying angles until a contrasting line pattern is detected, the position and alignment of the data field relative to the image window being deter-mined in a second step and, in a third step, a raster scan in the direction of the data track being performed and the signals contained on the data track being read and decoded, the apparatus having an optoelectronic scanning device including a rotatable scanning raster which emits at the output the binary video signal corresponding to the image field which is scanned line-by-line and includes binarily the contrast pattern of the scanned line, a decoder for identifying the contrasting line pattern which identifies the position and orientation of at least one data track of a data f ield, means for aligning the scanning raster parallel to the data track and for reading the scanned indicia of the data track, the improvement to the decoder comprising a counting circuit which recei~es the video signal and counts the length of successive overlapping intervals of the video signal, at least one reference table for receiving successively counted inter-val lengths in pairs via a gate circuit and for emitting a comparison signal having a first amplitude when the compared interval lengths have a given ratio which cor-responds to the ratio of the corresponding interval of the contrasting line pattern, and an evaluation circuit for generating an identification signal during a succession of a given number of comparison signals each having the first - 7c - 11755'71 amplitude and wherein the counting circuit includes a timing circuit for generating gate pulses which correspond to overlapping intervals of the video signal the pulse lengths of which is determined by successive rising slopes and by successive falling slopes of the video signal, respective counters activated by the gate pulses, the counters being jointly connected at their inputs with a sync generator, the counters generating a digital output defining the measured interval length, and means for applying the digital output to a reference table, each read-only memory includes selectively addressable dif-ferent expected fields for differing contrasting line patterns, and wherein the comparison signals associated with the different fields are selectively fed via a switch to an evaluation circuit wherein the reference table comprises a two-dimensional read-only memory, wherein possible discrete counting values of a first interval length are addressed to associated lines of the read-only memory, wherein possible discrete counting values of the subsequently counted interval lengths are addressed to associated columns of the read-only memory, and wherein an expected field within the memory encompasses the memory positions where the quotient of line address to column address falls within a given value range, so that a com-parison signal with a first amplitude is emitted when a memory position within the expected field is addressed, and a comparison signal with a second amplitude is emitted when a memory position outside the expected field is addressed, each addressable location of the read-only memories stores a n bit data word (n > 1), wherein the expected field for a m-th contrast line pattern (PICm, n > m > 1) is stored in the m-th bit of the n bit data words by making this bit "hi", wherein the read-only memories are coupled to a gating circuit which 35 emits an output signal LPICm ( m = 1, 2, 3) when the read-only memories are addressed and one of the scanned contrast line patterns is identified.

- 7d -Since different contrast line patterns can be identified with a decoding circuit, it is possible to associate a certain information content with applied contrast line patterns. For example, an applied contrast line pattern may include information about the form of the indicia pattern in the data tracks. For example, a particular contrast line pattern may always be used if the first data track includes but alpha indicia, for example the name of the object to be identified. Alternatively, individual contrast line patterns may be printed before, behind or below the data tracks to signalize that the data tracks have a predetermined corresponding form.
If a particular contrast line pattern is identified, a corresponding individual identification signal is emitted, and this identification signal can be used to start corres-ponding control functions. For example, the identification signal can activate an Alpha-character decoder and a numer-ical character decoder, successively.
When the scanning beam sweeps over a darkly colored area of the data field, the binary video signal has a first amplitude "Hi", and it has a second amplitude "Lo"
when the scanning beam sweeps over a light, signal-free area of the 8 1175S7i data field. The allvcation of the amplitudes Hi and Lo is arbitrary, and a different allocation of the two amplitudes to light and dark areas of the data field is possible.
The intervals of the video signal, the length of which is to be measured, preferably extend from one rising slope of the video signal to the next rising slope, as well as overlapping therewith from one falling slope of the video signal between the rising slopes to the next falling slope.
The third interval then extends from the second rising slope to a next-rising slope, and so forth. The fact that the intervals extend from a rising slope to a rising slope, or from a falling slope to a falling slope ensures that blurred edges of the contrast lines of the contrasting line pat-tern -- which generally extend in the same direction and in lS the same manner on all contrast lines -- do not materially influence the interval lengths. Conseguently the decoder can identify contrasting line patterns in accordance with the invention which were produced under variable printing conditions.
To enhance the reliability of the identification, contrasting line patterns having a signal-free lead zone of a given length can be employed. This lead zone is disposed ahead of the first line of the pattern. In such a case the measurement of the interval length is preferably initiated only when a signal-free lead zone of a predetermined length appears in the video dignal.
Measurement of the interval lengths is preferably terminated when a measured interval length exceeds a pre-determined maximum which e~uals the maximum interval length present in the contrasting line pattern. An ongoing measure-ment of the interval lengths is preferably also terminated when the ratio of two successively measured interval lengths falls outside the predetermined value range. In both cases, the object identifying process is brought to a halt at the earliest possible moment and the decoder is reset and ready for a new cycle.
The ongoing measurement of the interval lengths is preferably also terminated when a signal-free intermediate 9 117557i zone of a predetermined duration which equals the maximum distance between the lines of the line pattern i~ detected.
The decoding device is then reset and i~ ready for a new cycle.
ThP contrasting line pattern further preferably has a signaL-free trailing zone the length of which corresponds, for example, to the length of the signal-free lead zone. In such an event a contrasting line pattern identification signal is preferably emitted only when~-the video signal also includes ~he signal-free trailing zone of predetermined length.
, To enhance the redundancy of the identification process and thus reduce the probability of error, the same line is preferably scanned n times, and a contrasting line pattern identification signal is emitted only when the pat-tern has been successively identified n times.
The lengths of the overlapping intervals of the video signal are preferably digitally measured and for this purpose they are counted out in a counting circuit The counting circuit includes a timing circuit which generates gate pulses of the same length as the corresponding intervals of the video signal. The length of a given interval of the video signal is measured with a counter which receives as an input the gate pulses of that interval. The inputs of the counters are further connected with a sync generator which emits sync pulses to-the counters. The final count of the individual counters then corresponds to the length of the r~spective gate pulses and thereby to the length of the corresponding intervals. Successively counted interval values address the corresponding read-only memory after the subsequent interval value has been co-lnted, and while coun-ting of further interval values may still continue.
As already mentioned, each addressable location of the read-only memories stores a n-bit data word, n being preferably 8. The expected field for the m-th contrast line pattern PIC is stored in the m-th bit of the n-bit data words by keeping the respective bit "Hi". The outputs of the read-only memories are connected to a gating circuit, which emits an output after the addressing of the read-Gnly memories, said output specifying which of the possible contrast line lo il7~
patterns PICm is identified. If for example a first count is measured that addresses the line of the first read-only memory PROM 1, and if then a second count is measured which addresses the column of the first read-only memory PROM 1, the memory location defined by the addressed line and column emits a n-bit data word. The same applies for the second and third count which addresses a location in ~e second read-only memory, and the same applies for the third and fourth count which addresses a location in the third read-only memory.

The gating circuit connected to the outputs of the PROM's includes preferably n parallel AND-gates, n ~m ~1; m = 1, 2, 3 ..
The number of input terminals of the m-th AND-gate equals the number of comparison tables or expected fields, i.e. the number of read-only memories. The input terminals of the m-th AND-gate receive the m-th bit of the n-bit data words read from the different read-only memories.
The m-th AND-gate emits an output signal ~PICm, m = 1, 2, 3 ... when all its input terminals receive the value Hi, i.e. when the m-th bit from ~e different read-only memories signalize that the quotient of the successively measured counts compared in the respective read-only memory lies inside the m-th expected field.
Preferably the apparatus is organized such that an identification of the scanning direction relative to the line sequence of a contrast line pattern PICm is possible. For this purpose the successively measured interval length of the applied contrast line pattern PICm as scanned in forward direction (from left to right) is associated to a first expected field. The successively measured interval length of the same contrast line pattern PICm as scanned in reward direction (i.e. from right to left) is associated to a second expected field (comparison t~ble). When identifying the contrast line pattern in forward direction and in rean~ddirection, respectively, an identification signal PICm OUT V and PICm OUT R is emitted, respectively, thus indicating also the scanning direction. It is the advantage of this embodiment of the invention that the sequence, i.e. 1, 2, 3 of the contrast lines can be identified when scanning this sequence in the direction 1, 2, 3, and when scanning this sequence in oppositve direction, i.e. in the sequence 3, 2, 1.
4~ By this feature a field rotation of maximally 180 - in ~175571 different discrete angles - suffices to find any contrast line pattern, so that the searching time is reduced to about one half of the normal value.
In the following drawings Figures 1 - 12 correspond with those in the previous Wevelsiep application, while Figure 11 relates to an embodiment of the present invention.
Figure 1 shows a first arrangement of a contrasting line pattern within a data field having a data track Figure 2 shows a second arrangement of a contrasting line pattern within a data field;
Figure 3 shows a third arrangement of a contrasting line pattern within a data field;
Figure 4 shows the light-dark distribution of various contrasting line patterns taken perpendicular to the contrast lines;
Figure 5 shows a portion of the video signal as a function of time which corresponds to the contrasting line pattern of Figure 4(a);
Figure 6 is a block diagram of the counting circuit of the decoder;
Figure 7 is a block diagram of the comparison table of the decoder;
Figure 8 is another embodiment of the comparison table of the decoder;
Figure 9 is a block diagram of the evaluation circuit of the decoder;
Figure 10 is a schematic representation of the structure of the comparison table;
Figure 11 is a pulse diagram of the pulses generated in the timing circuit of the counting circuit;
Figure 12 is a pulse diagram or the pulses processed within a selecting circuit; and Figure 13 is a circuit diagram of a further embodiment of the comparison table according to Figure 8, including a gating circuit for selectively reading out different expected fields of each read-only memory.

117557i Figures 1 to 3 show a variety of object identifications 50 such as adhesive price labels which are secured, for example, to a container, a package or on any other article (not separately shown), and which appear in random positions and orientations on an image window. The image window is defined, for example, by the optical aperture of a flying-spot scanner such as a vidicon, which first scans the image window line-by-line, and then in a linewise raster scan.
The identifications 50 have a data field which includes contrasting signals 52 in at least one data tract 51 for identifying the object or article. The contrasting signals are preferably optical character signals of one of the known, machine readable types, for example OCR-A or OCR-B characters.

117~S'.Yl A contrasting line pattern 54 -- often referred to as position identification code or "PIC" -- is in a predeter-mined position and orientation in relation to the date track and has a plurality of parallel contrast lines with varying spacing and/or line widths. In the embodiment shown in Fig.
l, the line pattern is located in advance of the date track, in the embodiment shown in Fig. 2 it is underneath the data track, and in the embodiment shown in Fig. 3 it is at the end of the data track. The contrasting line pattern 54 is asym-metrical in a direction perpendicular to the contrast linesso as to identify the data field with regard to the beginning and the end of the data tracks. ~he line patterns shown in Figs. 1 and 2 have a signal-free lead zone 56 and a signal-free trailing zone 58.
Although the illustrated line patterns have only three lines each, patterns having more than three lines may be used. Further -- and deviating from the illustration of Figs. 1 to 3 -- the patterns may be located at a different position and have a different orientation in relation to the data tracks. It is further possible to provide two or more line patterns on one identification field 50.
A shown in Fig. l, the image window, or an image corresponding to the window, for example on the target of a vidicon, is scanned step-~y-step under an angle a by at least one scan line 60. Before reading the data tracks, it is important to first reliably identify the line pattern and determine its posi-tion and orientation relative to the scan line 60 of the scanning beam since the signals contained in the data track can then be read by subsequent raster scanning in the direction of the data track.
Fig. 4 shows the light-dark distribution of various three-line PIC patterns ta~en perpendicular to the direction of the individual lines which are all asymmetric and can therefore be used in accordance with the invention.
Fig. 5 shows a section of the video si~nal o~tained from scanning a PIC pattern in accordance with Fig. 4(a) as an electric binary signal, the amplitude Hi being allocated to the dark areas of the pattern and the amplitude Lo to the 1175~
` 14 light areas of the pattern. Light-dark fluctuations within the individual lines and the spacings of the PIC pattern are elimin~ted from the electric signal immediately after scan-ning. The video signal section shown in Fig. 5 includes a signal free lead zone that corresponds to the lead zone 56 in Fig. 1, a first interval Tl which extends from the first ascending flank or slope to the second ascending flank or slope, a second interval T2 which extends from the first descending flank to the second descending flank, a third interval ~3 which extends from the second ascending flank to the third ascending flank, and a fourth interval T4 which extends from the second descending flank to the third descending flank. It also includes a trailing zone that corresponds to the trailing zone 58 of Fig. 1.
The PIC pattern, for example the one shown in Fig.
~, is decoded according to the delta distance method, which determines whether successive and overlapping intervals, that is Tl, T2 and T2, T3 and T3, T4 have a predetermined ratio relative to each other as given by the PIC pattern which is to be decoded. If the value of the guotients of successive overlapping interval lengths falls within predetermined value ranges, the size of the range being determined by printing blurrs or digitalization inaccuracies, then in all probabil-ity the searched-for PIC pattern is present.
Fig. 6 shows the counting circuit, which forms the input of the decoder:of the invention, and which counts the interval lengths Tl to T4 and makes them available as binary values for further processing. The counting circuit contains a timing circuit 2 to which the video signal "VIDE0" is fed and which emits at a first output a first gate signal from a first rising slope to a second rising slope of the video signal, a second gate signal T2 at a second output from a falling slope following the first rising slope to a second falling slope of the video signal, a third gate signal T3 at a third output from the second rising slope to the next, third, rising slope, and a fourth gate signal T4 at a fourth output from the second falling slope to the next, third, falling slope. See also the pulse schematic shown in Fig.
11 .

11755'~1 The gate signals Tl and T4 are fed individually to the gate inputs Gl to G4 of the four counters 6, 8, 10 and 12, respectively. ~ach counter recei~es at its input CT1 to CT4 sync pulses from a sync generator 14 which are counted by S the counters so long as the respective gate signals Tl to T4 are applied. The result obtained at the outpus TCl to TC4 of the counters 6 to 12 then represents a measure for the length of the gate signals Tl to T4.
In the timing circuit 2, a release signal E2 is generated by the falling slope of the gate signal T2, a release signal E3 is generated by the falling slope of the gate si~nal T3, and a release signal E4 is generated by the falling slope of the gate signal T4. The release signals are emitted at separate outputs. The release signal E4 further lS generates a signal PWAIT the length of which corresponds to the signal-free trailing zone after the end of the gate signal T4, and it too is emitted at a separate output.
The counting circuit further includes a reset circuit 4 which receives the video signal VIDEO, and an external reset signal "RESET IN" at a separate input at the start of each scan line. The reset circuit 4 emits a reset signal RESET to the reset inputs RSl to RS4 of the counters 6 to 12 as well as to the timing circuit 2 and resets the counters 6 to 12 as well as the timing circuit 2 into an ( , 25 active starting condition when the video signal contains a signal-free section -- of an amplitude Lo -- which is larger than the maximum scanning distance between the lines of the PIC patterns as given ~y the maximum spacing within the PIC
pattern, multiplied by the largest permissible scanning angle.
The counting circuit further contains an overflow sensor 40 which is responsi~e to an overflow or carryover output OV1 to OV4 of the counters 6 to 12 and emits a reset signal "oV RESET", and then resets the decoder into a new state of readiness.
Fig. 7 shows an embodiment of the comparison table of the decoder which includes a read-only memory or PROM 28.
PROM 2~ is organized so that count TCl of the counter 6 16 117~

addresses individual lines of the memory matrix and count TC2 of the counter 8 addresces individual columns of the first memory matrix. Counts TCl and TC2 are applied to PROM 28 by t~e release signal E2 via the gate circuits 16, 18 after TC2 has been counted. The first memory matrix includes an expec-tancy field which encompasses the memory positions where the guotient of the line address and the column address falls within a predetermined value range. This value range cor-responds to the guotient of a first interval length to a second in~terval length of the PIC pattern under considera-tion. If a memory position within the expectancy field is addressed by counts TCl and-~TC2, a reference signal having a first amplitude, for example Hi, is emitted which signals that information had been scanned which corresponds to a portion of the PIC pattern under consideration.
To correspondingly compare the gate pulses T2 with gate pulse T3, the count TC2 of the second counter 8 addres-ses the lines of a second memory matrix, and the count TC3 of the third counter 10 addresses the columns of the second-memory matrix. The addressing takes place via gates 20, 22after the count TC3 has been counted and the counts TC2 and TC3 are sent to PROM 28 ~y the release signal E3. The second matrix also includes an expectancy field which encompasses the memory positions whose guotient of line address and column address falls within a predetermined range which is equal to the value range of the guotient of the intervals of the PIC pattern corresponding to gate signals T2 and T3.
When a memory position in the expectancy field is addressed, ; a reference signal having a first amplitude, for example the amplitude Hi, is emitted. When a memory position outside the expected field is addressed, a reference signal having a second amplitude Lo is emitted.
Comparison of the count TC3 of the third counter 10 with the position TC4 is accomplished likewise by addressing lines and columns of a third memory matrix which also includes an expectancy field. When a memory position within the expectancy field is addressed, a reference signal having a first amplitude Hi is emitted. The third memory matrix is li7~571 addressed via gates 24, 26 after the count TC 4 of the fourth counter 12 has been counted and the counts TC3 and TC4 are sent t~ the PROM 28 by the release signal E4. Gates 16 to 26 comprise AND gates.
The release of comparison signals LPIC1 and LPIC2 and LPIC3 is effected by the release signals EL2, EL3 and EL
4 which are o~tained by delaying the release signals E2, E3, and E4 in the delay circuit 30; see also the pulse plan of Fig. 12. Read-out may only take place after the first, second and third memory matrix have been addressed.
As an alternative to the embodiment shown in Fig.
7, the first memory matrix may be defined by a first read-only memory, PROM1, the second memory matrix by a second read-only memory, PROM2, and the third memory matrix by a third read-only memory, PROM3. In this embodiment three read-only memories of relatively low storage capacity can be employed.
The read-only memory 28 has storage locations each of which has an n-bit capacity, and for example n = 8. Since for the provision of an expected field only one bit of each storage location defining the expected field is occupied, up to n different expected fields for n different PIC patterns can be simultaneously accommodated, whereby preferably n = 8, and the first expected field in PROMl is accommodated in the first bit of the memory positions, the second expected field in the second bit of the memory positions, and so forth. The same applies for PROM2 and PROM3. The line and column addressing for a specific PIC pattern must then occur selectively to the corresponding bits of the memory positions. Further, a switch 32 coupled to PROM28 selectively reads out the com-parison signals LPIC1 and LPIC2 and LPIC3 from the pertinent bits of the memory positions and transmits as its output to an evaluation circuit the evaluation signal MUX PIC, formed of the sequential comparison signals LPICl, LPIC2 and LPIC3.
Fig. 9 illustrates the evaluation circuit of the decoding device. An interim memory 34 receives the evalua-tion signal MUX PIC and stores the comparison signal LPICl --which indicates that the value TCl/TC2 falls within a pre-' ~ 1175571 determined range -- as well as comparison signals LPlC2 and LPIC3. Storing is commenced by release signals EL2 and EL3 and EL4 which are qenerated substantially simultaneously with the comparison signals LPICl, LPIC2 and LPIC3, see the pulse schematic of Fig. 12. After all comparison signals have been stored in the interim memory as storage signals LPICl', ~PIC2', LPIC3', the storage signals are transmitted to an AND
gate 36 which emits an output signal LPIC when all storage signals LPICl' etc. have a first amplitude corresponding to the first-amplitude of the comparison signals LPICl etc., see Fig. 12. The output signal LPIC is fed to an output circuit 38 which receives the video signal VIDEO and the hold signal PWAIT from the control circuit 2. The output circuit 38 generates an identification signal "PIC OUT" when the video signal VIDEO remains on amplitude Lo while hold signal PWAIT
is applied. The amplitude Lo identifies a signal-free sub-surface. This ensures that the decoded line pattern is followed by a signal-free trailing zone which corresponds to the trailing zone ~ of the PIC pattern.
The output circuit 38 is reset by the external reset signal "RESET IN" and thereupon applies a reset signal RESETA to the interim memory 34 and resets the latter for a new cycle. The interim memory is further reset by the over-flow reset signal "OV RESET" when one of the counters 6 to 12 signals an overflow.
Fig. 10 is-a schematic representation of the organ-ization of the comparison table, for example the partial comparison ta~le of PROMl for comparing the guotient TCl/TC2.
The table comprises a memory matrix and its lines and columns have the appropriate binary addresses. In accordance with a preferred embodiment of the decoder o the invention, a 5-bit representation has been selected. All memory positions with a specific value of the guotient of line address to column address lie on one line, the so-called expectancy line around which the expectancy field is located. Within the field all those memory positions are located which address quotients that fall in the predetermined value range. The counts TCl to TC4 are also emitted as 5-bit words. The ount TCl , g 1175~'7~

addre~ses the lines of the table, the count TC2 addresses the columns of the table.
Fig. 13 shows a circuit diagram of a further embodiment of the comparison table according to fig. 8. The comparison table S includes a first read-only memory PROM 1, the lines of which are addressed by the measured first count TC1, and the columns of which are addressed by the second measured count TC2,included is further a second read-only memory PROM 2, and the measured second count TC2 addresses its lines, the measured third count TC3 addresses its columns. Included is additionally a third read-only memory PROM 3, and the measured third count TC 3 ; addresses its lines, the measured fourth count TC 4 addresses its columns. Thus, each read-only memory PROM1, PROM 2, ...
provides a two-dimensional line and column organized storage matrix to realize comparison tables for respective two successively measured counts of the contrast line pattern.
Each location of the memories PROM 1, PROM 2, ...
addressed by its line and column, stores a n-bit data word with n 1, preferably n = 8. For identifying m different contrast line patterns, with n ~, m 3 1, the expected field for a m-th contrast line pattern PICm, m = 1, 2, 3 ..., is stored in the m-th bit of the n-bit data words by writing the value Hi in the respective bit whereas the n-th bitsof memory locations outside the expected field have the amplitude Lo.
The output of the read-only memories PROM 1, PROM 2, ... is fed to a gating circuit 28, with n parallel AND-gates 29. The AND-gates 29 have as many input terminals asexist read-only memories PROM 1, PROM 2, ...
Between the output of each read-only memory PROM 1, PROM 2, ... and the gating circuit there are provided interfaces 29a, which emit at their output the data word received from the read-only memory in bit parallel form. The interfaces 29a have n outputs, the first output emitting the first bit, the second output the second bit, the third output the third bit etc. of the received n-bit data word. The n AND-gate 29 are coupled to the interfaces 29a such th~t the m-th AND-gate receives at its different input terminals as input signal the m-th bit of the data words received from the different read-only memories. Each AND-gate emits an output signal LPICm when all its input signals have the amplitude Hi, i.e. when the m-th bit of all read-only memorieshave the amplitude Hi, thus, signalizing that the actually read contrast line pattern PICm lies in the m-th expected field and is identified without failure.
Thus, when the first AND-gate 29 emits a pulse, a first contrast line pattern PIC1 is read and identified. If the m-th AND-gate 29 emits a pulse, the m-th contrast line PICm is read and identified.
The output of the n AND-gates are connected to a selection circuit (not shown). This selection circuit is adjustable such that it emits an output signal only when the select~d contrast line pattern PIC is read and identified.

Claims (6)

Claims:

1. In a method for identifying objects appearing at random positions in random orientation, and at random times on an image window and having, on a surface facing the image window, an identification in the form of a field which includes on at least one data track contrasting indicia with at least one contrasting line pattern iden-tifying the position and the orientation of the data track and having a plurality of parallel lines with variable spacing and/or line widths, the image window being scanned line-by-line and a binary video signal being generated which corresponds to the scanned contrast sequence, the image window being, in a first method step, scanned under varying angles until a contrasting line pattern is detected, the position and alignment of the data field relative to the image window being determined in a second method step and, in a third method step, a raster scan in the direction of the data track being performed and the indicia present on the data track being read and decoded, the contrasting line pattern being identified by measuring the length of the overlapping light-dark intervals of the video signal, comparing the successively measured interval lengths, generating a comparison signal having a first amplitude when two interval lengths which are being compared have a predetermined ratio to each other which corresponds to a corresponding spacing in the contrast-ing line pattern, and emitting an identification signal when, during each of a number of successive comparing steps as determined by the contrasting line pattern, the comparison signal is received, the comparing of successively measured interval lengths is carried out in pairs in a two-dimensional comparison table within which the possible discrete values of an interval length are assigned to table lines, and the possible discrete values of the next following measured interval length are assigned to table columns, and wherein an expected field is specified encompassing the positions on the table in which the quotient of the two successively measured compared interval lengths falls within a given value range, and including the step of generating the refer-ence having a first amplitude when the compared interval lengths correspond to a table position within the expected field, differing expected fields for corresponding success-ively measured interval lengths of various contrasting line patterns are included in the individual comparison table, characterized in that an individual identification signal for each contrasting line pattern is emitted when identifying the corresponding contrasting line pattern.

2. A method according to claim 1 wherein each individual comparison table includes a first expected field corres-ponding to successively in forward direction measured interval length of a contrast line pattern, and a second expected field corresponding to successively in rearward direction measured interval lengths of this contrast line pattern, and wherein a first identification signal (PIC OUT V) is emitted when the contrast line pattern is identified by scanning in forward direction, and wherein a second identification signal (PIC OUT R) is emitted when the contrast line pattern is identified by scanning in rearward direction.

3. A method according to claim 2 wherein each iden-tification appearing on the image window comprises a plurality of different contrast line patterns (PICm, m = 1, 2, 3, ...) and wherein an individual identification signal (PICm OUT m = 1, 2, 3, ...) is emitted when the corresponding contrast line pattern is identified in forward and rearward direction, respectively.

4. In an apparatus for identifying objects appearing at random positions in random orientation, and at random times on an image window and having, on a surface facing the image window, an identification in the form of an image field which includes on at least one data track contrasting indicia with at least one contrasting line pattern identifying the position and the orientation of the data track and having a plurality of parallel lines with variable spacing and/or line widths, the image window being scanned line-by-line and a binary video signal being generated which corresponds to the scanned contrast sequence, the image window being, in a first step, scanned under varying angles until a contrasting line pattern is detected, the position and alignment of the data field relative to the image window being deter-mined in a second step and, in a third step, a raster scan in the direction of the data track being performed and the signals contained on the data track being read and decoded, the apparatus having an optoelectronic scanning device including a rotatable scanning raster which emits at the output the binary video signal corresponding to the image field which is scanned line-by-line and includes binarily the contrast pattern of the scanned line, a decoder for identifying the contrasting line pattern which identifies the position and orientation of at least one data track of a data field, means for aligning the scanning raster parallel to the data track and for reading the scanned indicia of the data track, the improvement to the decoder comprising a counting circuit which receives the video signal and counts the length of successive overlapping intervals of the video signal, at least one reference table for receiving successively counted inter-val lengths in pairs via a gate circuit and for emitting a comparison signal having a first amplitude when the compared interval lengths have a given ratio which cor-responds to the ratio of the corresponding interval of the contrasting line pattern, and an evaluation circuit for generating an identification signal during a succession of a given number of comparison signals each having the first amplitude and wherein the counting circuit includes a timing circuit for generating gate pulses which correspond to overlapping intervals of the video signal the pulse lengths of which is determined by successive rising slopes and by successive falling slopes of the video signal, respective counters activated by the gate pulses, the counters being jointly connected at their inputs with a sync generator, the counters generating a digital output defining the measured interval length, and means for applying the digital output to a reference table, each read-only memory includes selectively addressable dif-ferent expected fields for differing contrasting line patterns, and wherein the comparison signals associated with the different fields are selectively fed via a switch to an evaluation circuit wherein the reference table comprises a two-dimensional read-only memory, wherein possible discrete counting values of a first interval length are addressed to associated lines of the read-only memory, wherein possible discrete counting values of the subsequently counted interval lengths are addressed to associated columns of the read-only memory, and wherein an expected field within the memory encompasses the memory positions where the quotient of line address to column address falls within a given value range, so that a com-parison signal with a first amplitude is emitted when a memory position within the expected field is addressed, and a comparison signal with a second amplitude is emitted when a memory position outside the expected field is addressed, each addressable location of the read-only memories stores a n bit data word (n ? 1), wherein the expected field for a m-th contrast line pattern (PICm, n ? m ? 1) is stored in the m-th bit of the n bit data words by making this bit "hi", wherein the read-only memories are coupled to a gating circuit which emits an output signal LPICm ( m = 1, 2, 3...) when the read-only memories are addressed and one of the scanned contrast line patterns is identified.

5. Apparatus according to claim 4, wherein the gating circuit includes n parallel AND-gates, n ? m ? l;
m = 1, 2, 3 ..., wherein the m-th bit of the data words stored in the read-only memories are coupled to the inputs of the m-th AND-gate which emits the output signal LPICm when all inputs are "hi".

6. Apparatus according to claim 4 or claim 5, wherein a first expected field is provided for the successively measured intervals counts of a contrast line pattern scanned in forward direction (for example scanned from left to right), wherein a second expected field is pro-vided for the successively measured interval counts of said contrast line pattern scanned in rearward direction (scanned from right to left), and that an identification signal is emitted when identifying the contrast line pattern, said identification signal including information which characterizes the relative scanning direction.